Topology and porosity modulation of polyurea films using interfacial polymerization

Abstract

Polyurea (PU) films are interesting owing to the high mechanical properties and wide range of applications. Here we report an interfacial synthesis of a PU thin film at the liquid–liquid interface using the reaction between diisocyanate and polyamines. Three amines – polyethylenimine (PEI), diethylenetriamine (DETA) and tris(2-aminoethyl)amine (TREN) were dissolved in the aqueous phase separately and allowed to react with an organic solution of hexamethylene diisocyanate (HMDI) at the liquid–liquid interface. The structure and morphologies of the films were established using spectroscopic and microscopic techniques. Smooth PU films were obtained from the reaction of HMDI with PEI, whereas films with holes and tubules were obtained from small molecular amines such as DETA and TREN. Such observations are explained based on the differences in solubility, reactivity of the reagents and molecular size dependant diffusion across the film. The effect of the changes in concentration of reactants and time of reaction on film thickness and transport of organic dye molecules through the film were investigated. PEI gave a smooth defect-free film of PU, whereas small molecular amines upon reaction with HMDI gave films with significant number of defects. To demonstrate the accessibility of functional groups on the film, adsorption of fluorescent dyes on the film surface was investigated. From the UV-Vis spectroscopic measurements, it is clear that acidic molecules adsorb on the surface of the films and the corresponding esters are not extracted efficiently by the film. In summary, the synthesis and characterization of PU films from multiple amines, and the transport of small molecules and adsorption of dyes on the surface of the films were established.

---

Introduction

Polymer films are employed in coatings,¹ separation,² nanofiltration^2b,3 and in organic electronics.⁴ Polyamides, polyurethanes and polyureas (PU) form high quality films with high mechanical and thermal stability, high flux and good selectivity of molecules or salt rejection properties.⁵ Different methods including molecular layer deposition (MLD),⁶ atomic layer deposition (ALD),⁷ layer-by-layer assembly (LbL),⁸ chemical vapor deposition (CVD)⁹ and spin coating techniques⁴ have been employed for the formation of ultrathin polymer films on different substrates. As an alternative, self-assembly and polymerizations at the oil–water interface have received extensive interests in the formation of thin films owing to the ease of synthesis and lower number of steps involved.¹⁰ Interfacial polycondensation (IP) involves reaction of two or more reagents at the liquid–liquid interface of two immiscible solvents. Such reaction may involve the synthesis of small molecules, polymers or cross-linked polymer networks.^2c,11 IP is also used in the synthesis of microcapsules (MC),¹² bulk polymers,¹³ polymer nanocomposites¹⁴ and self-healing materials.¹⁵ It is possible to structure the polymeric material during IP by controlling chemical composition, concentration of monomers, pH and reaction time. Furthermore, no purification steps are necessary and the films can be easily transferred to a substrate.

Polyureas (PU) are known to exhibit interesting mechanical, thermal and chemical stability, abrasion resistance, water repellency, high rate of healing and viscoelastic properties.¹⁶ Thin films and microcapsules of PU are prepared via condensation polymerization of nucleophilic polyamines and reactive diisocyanates to form urea linkages under different conditions.¹⁷ Recently, microfluidic devices were used to produce monodispersed PU microcapsules with controlled thickness.¹⁸ In addition, theoretical modeling of microencapsulation process and the formation kinetics of PU shells are also reported.^11c,19 Furthermore, there are a few reports on preparation of PU films via IP. PU film formation on the Teflon support in water–cyclohexane interface,^11b PU–amide membranes for RO/UF application²⁰ and PU macroporous film formation at the interface of hexane and ionic liquid were reported earlier.²¹

Herein, we report the synthesis, full characterization of structure, surface morphology and reactivity of surface functional groups of PU films prepared using liquid–liquid interface polymerization (IP). Aqueous solutions of polyethyleneimine (PEI), diethylenetriamine (DETA) and tris(2-aminoethyl)amine (TREN) were reacted with HMDI dissolved in organic solvents (chloroform, carbon tetrachloride or toluene) to form PU films. PEI was chosen as the polymeric reagent with high amine density but low diffusion rate; DETA and TREN were selected as smaller amine monomers which could show different reaction rate and surface morphologies. The objective of the current study includes developing two-dimensional PU film, optimizing experimental factors, investigating the structure–property relationship of the films formed from polymeric reactant and small molecular amines and to demonstrate potential application as adsorbent for organic dyes.

Experimental

Materials and chemicals

Hexamethylene diisocyanate (HMDI), diethylenetriamine (DETA), 1-pyrenecarboxylic acid (1-PCA) and 9-anthracenecarboxylic acid (9-ACA) were obtained from Sigma Aldrich with high purity. Polyethylenimine (PEI), M_n = 60k was obtained as a 50% aqueous solution from Sigma Aldrich. Tris(2-aminoethyl)amine (TREN) was purchased from Fluka Chemicals. All chemicals were of AR grade purity and used without further purification. AR grade toluene, carbon tetrachloride and chloroform were supplied by VWR Chemicals Singapore. Deionised water was used to dissolve amines in all experiments. Ethyl esters of 1-PCA (PCEt) and 9-ACA (ACEt) were synthesized via acid chloride formation using literature report.²²

Preparation of polyurea thin film by reacting PEI with HMDI

The PU films were prepared by taking solutions of reactants in two different immiscible solvents and allowing the reaction to take place at the liquid–liquid interface (Scheme 1). In a typical experiment, to a solution of HMDI in CHCl₃ (1 mg mL⁻¹, 0.006 mmol mL⁻¹, 10 mL) placed in a 25 mL beaker, aqueous solution of PEI (1 mg mL⁻¹, 0.016 mmol mL⁻¹, 10 mL) was added slowly and allowed the reaction to proceed at the interface at room temperature for 4 h. After the film formation, top aqueous layer was carefully removed using a dropper. The exposed film was transferred onto a glass plate, rinsed with CHCl₃, followed by water to remove residual starting materials and dried in open air for further characterization. Experiments were also carried out with aqueous PEI solution (1 mg mL⁻¹, 10 mL) and a solution (10 mL) of HMDI in CHCl₃ with varying concentrations – 5 mg mL⁻¹ (0.03 mmol mL⁻¹), 3 mg mL⁻¹ (0.018 mmol mL⁻¹) and 0.5 mg mL⁻¹ (0.003 mmol mL⁻¹). A control experiment was performed with HMDI in CHCl₃ layer and water in the top layer. No film was formed after 12 hours in the absence of amines in the aqueous phase.

Scheme 1 General experimental setup and the reaction scheme for the formation of PU film at the liquid–liquid interface.

To check the effect of organic solvents, the experiment was repeated by dissolving HMDI in toluene and reacting with aqueous solution of PEI as discussed above. The concentration of HMDI was varied (0.003, 0.018, 0.03 mmol mL⁻¹), while maintaining a PEI concentration at 1 mg mL⁻¹ (0.016 mmol mL⁻¹).

Preparation of films from reaction of TREN or DETA with HMDI

Aqueous solution of DETA (5 mg mL⁻¹, 0.048 mmol mL⁻¹, 10 mL) and a solution of HMDI (8.2 mg mL⁻¹, 0.048 mmol mL⁻¹) in toluene (10 mL) were taken in a 50 mL beaker. After 4 h, the film formed at the liquid–liquid interface was transferred carefully on a glass plate, washed with water, dried and used for further characterizations. Similar PU films were formed at the interface of 10 mL aqueous DETA solution (5 mg mL⁻¹, 0.048 mmol mL⁻¹) and CHCl₃ solution of HMDI at various concentrations – 16 mg mL⁻¹ (0.096 mmol mL⁻¹), 25 mg mL⁻¹ (0.144 mmol mL⁻¹), 33 mg mL⁻¹ (0.192 mmol mL⁻¹) and 41 mg mL⁻¹ (0.24 mmol mL⁻¹). The film morphology and thickness were measured using FESEM analyses. Experiments were repeated by taking aqueous solution of TREN (5 mg mL⁻¹, 0.034 mmol mL⁻¹, 10 mL) and toluene solution of HMDI (5.75 mg mL⁻¹, 0.034 mmol mL⁻¹, 10 mL) in a beaker, isolating and characterizing the film formed at the interface after 4 hours. No precipitates were observed in either phase, implying the reaction only occurred at the interface.

In order to check the influence of solvents, HMDI dissolved in CCl₄ (Warning: Toxic chemical and must be used inside the fume hood) was used for the reaction with DETA in aqueous solution (5 mg mL⁻¹, 0.048 mmol mL⁻¹). HMDI concentrations were in different mole ratios from 1 to 5 with respect to the amines and the thickness of the film obtained was monitored using FESEM analysis.

Adsorption experiment with fluorescent dyes

9-Anthracenecarboxylic acid (9-ACA) and 1-pyrenecarboxylic acid (1-PCA) were chosen as fluorescent dyes for adsorption on the film (Scheme 2). PU film formed from PEI/HMDI at the CHCl₃/water interface was exposed to the fluorescent compounds. In a typical procedure, a solution of 9-ACA (2 mg, 9 μmol) in CHCl₃ (1 mL) was injected slowly into the organic layer, with simultaneous removal of 1 mL solution from the organic phase from the opposite side of the container, which helps to reduce the pressure on the PU film. The PU film was collected on a quartz plate after an hour and washed with CHCl₃ to remove excess dyes. The emission of dyes adsorbed on the PU film was recorded using a UV spectrophotometer. Similar experiments with the corresponding ethyl ester derivatives did not show any significant adsorption. No transport of molecules across the PU membrane was observed owing to the poor solubility of such molecules in aqueous solution.

Scheme 2 Fluorescent molecules used for the adsorption studies.

Studies on the permeability across the film membrane

To analyze the permeability of the PU film, phenolphthalein was chosen as the dye which is a well-known pH indicator used in acid–base titrations. In acidic or neutral medium, it exists in lactone form and is colorless, whereas, phenolate anions are formed in basic medium (pH > 8.2), turning the solution pink. In a typical procedure, PU film was prepared from PEI/HMDI at the interface according to the procedure given above. Appropriate volume of phenolphthalein solution (3 mL, 9 μmol) in CHCl₃ was injected to the organic layer slowly using a syringe through the side of the plastic beaker. An equal volume of CHCl₃ fraction was removed simultaneously from the opposite side to avoid any rupture or damage of the PU film during the addition of phenolphthalein in the organic phase. In the case of PEI/HMDI, the experiment was repeated with different PU films with an average thickness of 80 nm, 200 nm and 290 nm. Aliquots of aqueous solution were analyzed periodically using a UV-Vis spectrophotometer to determine the amount of phenolphthalein transported across the membrane. It is conceivable that the rate of permeation of molecules depends on the thickness and porosity of the film.

Analyses and characterization

The PU films obtained were characterized using different techniques. FT-IR spectroscopy was performed on Bruker ALPHA FT-IR spectrophotometer in the spectral range of 500–4000 cm⁻¹. The film was dried under vacuum prior to the analysis and IR spectra were recorded using KBr as matrix. Surface morphologies and film thickness were determined using Field Emission Scanning Electron Microscope (FESEM on a JEOL JSM-6701F instrument) at 5 kV after coating the films with platinum. Atomic Force Microscopy (AFM) was performed using Agilent AFM with Pico plus molecular imaging system in non-contact tapping mode. X-Ray diffraction (XRD) was recorded using Bruker – AXS: D8 DISCOVER with GADDS Powder X-ray diffractometer with Cu-Kα (λ = 1.54 Å) at 40 kV and 40 mA within a range of 5° to 40° using a step size of 1°. The film was dried at 70 °C for 24 h and powdered well before recording XRD pattern. Thermogravimetric analyses (TGA) were conducted using a SDT 2960 TA instrument. All samples were heated under nitrogen atmosphere from 25 to 700 °C using a heating rate of 10 °C min⁻¹. The UV-Vis and emission spectra of the dye adsorbed films were measured on UV-1601PC Shimadzu spectrophotometer and RF-5301PC Shimadzu spectrophotometers, respectively.

Results and discussion

Formation of thin film at the liquid–liquid interface

Interfacial polymerization (IP) is used for structuring film at the liquid–liquid interface with preferred thickness and dimensions. The free-standing polyurea (PU) film isolated via the reaction of polyamine with diisocyanate is stable and can be used for various applications. Properties such as thickness can be controlled via optimization of structure and concentration of reagents (Fig. S1, ESI†). IP is affected by the concentration and partition coefficients of the reactants at the interface of two immiscible liquids, time of reaction and diffusion of monomers through the interface.²³ The rate of formation of PEI/HMDI film was examined by determining film thickness using SEM analysis after varying the concentrations of reactants and using a constant reaction time. The films formed after a fixed time interval were collected, washed and dried before SEM analysis. The effect of chemical nature of solvents on film thickness was measured by preparing films at the interface of two different solvents – CHCl₃/water and toluene/water. Fig. 1A shows the plot showing variation of film thickness after 4 hours with change in concentration of one of the reactants (HMDI) and the effect of solvents. Thickness of film increased linearly with increasing molar concentrations of HMDI dissolved in chloroform. However, the increase in thickness of film at the toluene/water interface was slow owing to the low solubility and low diffusion or partitioning of the amines into toluene.

Fig. 1 Effect of monomer mole ratio on the PU film thickness. (A) PEI/HMDI film growth at the interface of CHCl₃/water and toluene/water (PEI concentration of 1 mg mL⁻¹ and mole ratio (R) of HMDI to PEI is varied from 0.2 to 1.8). (B) Changes in DETA/HMDI film thickness at the interface of CCl₄/water and toluene/water (DETA concentration of 5 mg mL⁻¹ and R of HMDI to DETA is varied from 1 to 5).

The reaction of DETA or TREN with HMDI was performed at the toluene/water and CCl₄/water interfaces. Chloroform was not used owing to high solubility of the small molecular amines. The formation of thin film for DETA or TREN with HMDI was faster compared to PEI, because of the higher diffusion rate of smaller amine molecules through the membrane. Similar to the experiment with PEI/HMDI, a linear relationship of the effect of mole ratio of the reactants on the film thickness in the case of small molecular amines was established (Fig. 1B). In addition, polarity of the solvents also influences film characteristics. It is known that organic solvent used for dissolving HMDI influenced the kinetics of film formation with the reaction rate changing in the order cyclohexane > carbon tetrachloride > toluene.²⁴ A similar trend was observed in our studies with increase in film thickness was slow in toluene and a linear relationship between the thickness and mole ratio of reactants was observed in CHCl₃ and CCl₄ (Fig. 1).

Surface morphologies

The surface morphology is used to deduce information on the film growth.^11b,25 Films formed from the reaction of different amines with HMDI showed interesting surface morphologies. The SEM images of the PU film formed with PEI/HMDI at a concentration of 1 mg mL⁻¹ in CHCl₃/H₂O interface are shown in Fig. 2A and B. Film morphology on the organic solvent side (Fig. 2A) is relatively flat and smooth as compared to the porous network formation on water phase (Fig. 2B).

Fig. 2 FESEM micrographs of PU film with PEI/HMDI (A and B) at CHCl₃/H₂O interface at a concentration of 1 mg mL⁻¹ each, DETA/HMDI (C and D) and TREN/HMDI (E and F) with a mole ratio of 1 at toluene/water interface. Organic (A) and aqueous (B) sides of PEI film show rough surface after 30 minutes. Organic sides of DETA (C) and TREN (E) films show small tubes and spherical assemblies after 30 minutes and the corresponding film surface on the aqueous phase (D and F) show holes. After 24 hours, prominent long tubular structures (G) are seen on the organic side and large holes on the aqueous side (H) of DETA/HMDI film. Insets in (A–F) show magnified images.

The films obtained from the reaction of DETA or TREN with HMDI showed smooth surface on the water side and long structures with doughnut shaped ends and stem sizes ranging from nanometers to micrometers on the side of organic phase. The film obtained after 24 h of polymerization led to the formation of long tubules on organic side (Fig. 2G) and holes on aqueous side (Fig. 2H) of the film. Additional FESEM images of the PU films from the reaction of DETA or TREN with HMDI are shown in the ESI (Fig. S3, ESI†). Although, these structures appear to be tubular based on the SEM images, additional data is needed to confirm the exact structure of such overgrowth on the films. Interestingly, the aqueous side of the film showed holes and cavities with diameters comparable to the tube-type structures on the organic side (Fig. 2D).

Such growth is only seen in the case of PU films formed from DETA and TREN. The growth on the organic side is due to the continuous diffusion of small molecular amines through the film, which then reacts with diisocyanates and form tube-type structures. Such continuous diffusion of small molecular amines to the organic layer is possible owing to the partial solubility of amines in organic solvents. Since the diisocyanates are insoluble in water, no polymer growth was observed on the aqueous side. Since, PEI is insoluble in chloroform and expected to have slow diffusion rate, no such surface growths were seen in the case of PU films prepared from PEI. The mathematical models developed for IP of polyamide films included diffusion of amines through the membrane to react with other reagents in organic phase and the rate of the reaction at the liquid–liquid interface is limited by diffusion.²⁶

Polymerization, followed by deposition of materials led to the formation of PU film. In addition, the permeability of the film decreases with increase in thickness with time.²⁷ The diffusion-controlled process is a self-limiting process, where reagents diffuse and react with each other at the interface. The surface morphologies observed with DETA at different time intervals confirm such a process (see Fig. S2, ESI†). In the early stages, the bifunctional reagents react at the interface forming oligomers and polymers, which translate into a continuous film with increase in time. A thin layer of film was formed within 5 minutes and the tubule-type structures begin to appear on the organic side within 30 minutes. When the film formation was prolonged for 24 h, long tubular structures with lengths of a few micrometers were observed (Fig. 2G) and the aqueous side of the film showed corresponding holes.

Such tubular structures were not formed with PEI/HMDI film even after polymerization for a long duration of 24 h. The reactivity of a polyamine (PEI) with an average molecular weight of 60 000 Da with large number of reactive sites (e.g. –NH₂) is higher than the smaller amines – DETA and TREN – with a molar mass of 103.2 g mol⁻¹ and 146.2 g mol⁻¹. Due to the branched structure and multiple amine groups, the initial crosslinking reactions of PEI with diisocyanates lead to an extensive network structure. The diffusion of PEI molecule from water to CHCl₃ is limited due to its large molecular size and low solubility in organic phase. Smaller amines, in contrast, are partially soluble in chlorinated organic solvents and pass through defects in the film.

In addition to the SEM analyses (Fig. 2 and S3, ESI†), AFM height images (Fig. 3) are given over a scan area of 10 × 10 μm to visualize surface morphologies and measure surface roughness. PEI/HMDI film showed a rough morphology with a height profile of 50–60 nm (Fig. 3A) and DETA film showed spherical growth emanating from the surface with an average height of 500–1500 nm at the organic side (Fig. 3B). Holes or cavities were observed on the aqueous side of the DETA/HMDI film (Fig. 3C).

Fig. 3 AFM tapping mode images of organic side of PEI/HMDI (A) film prepared at CHCl₃/H₂O interface; organic (B) and aqueous (C) sides of DETA/HMDI film at toluene/H₂O interface. Insets show height profiles along the points, a → b.

For comparison of the morphologies polymers obtained via bulk polymerization, PU was synthesized using dropwise addition of HMDI solution (500 mg, 2.97 mmol) in THF (5 mL) to an equimolar solution of DETA (300 mg, 2.97 mmol) in THF (5 mL). The solid formed was washed with THF and dried. As expected the surface morphology of this bulk polymer is different from the films obtained using IP (Fig. S1C, ESI†).

FTIR analysis

Fig. 4 shows the FTIR spectra of the PU films in comparison with HMDI. The broad peaks centered at 3370–3420 cm⁻¹ correspond to hydrogen bonded N–H stretching. The –NCO peak of HMDI at 2287 cm⁻¹ was absent in PU films. Peaks at 2923 and 2853 cm⁻¹ correspond to the –CH– stretching vibrations of aliphatic chains. The formation of amide bonds is confirmed from the appearance of strong stretching vibration at 1633 cm⁻¹ (>C=O) and 1560 cm⁻¹ of amide II peak, mainly from in-plane N–H bending.

Fig. 4 FTIR spectra of PU films in comparison with HMDI. The spectra were recorded using KBr matrix.

XRD analysis

Broad X-ray diffraction (XRD) data indicate amorphous nature of PU film (Fig. 5A). Formation of crystalline domains in polymers is influenced by preparation and processing conditions of the polymer.^19a In particular, a higher rate of polymerization would lead to a lower degree of crystallinity, as the polymer grows faster than the chains arrange in crystalline order.^19b At the interface between two liquids, large surface area is available with 2D characteristics for the reaction between two monomers to form an instantaneous crosslinked and continuous polymer film. Under fast PU formation, polymer chains randomly organize into an amorphous lattice. This is evident from the XRD and DSC data collected from the PU film. The amorphous nature of PU film could also be the outcome of extended crosslinking inside the lattice.

Fig. 5 XRD (A) and TGA (B) traces of the PU films formed from HMDI and different amines.

Thermogravimetric analysis

Fig. 5B shows the thermal degradation curves of the PU samples formed from the reaction of HMDI with polymeric PEI or oligoamines. The initial weight loss at 100 °C for all three films is due to loss of solvent molecules trapped within the films. Since the films were formed at the interface between organic and aqueous media, incorporation of solvent molecules inside the PU film could be stabilized via H-bonds to the urea groups and complete removal of solvent molecules via drying before the TGA analysis was not achieved. The degradation of PU from DETA/HMDI followed similar pattern owing to analogous chemical structures and showed sharp degradation at 320–330 °C, whereas degradation started at 280 °C for PU from PEI/HMDI. While the early degradation of the latter can be attributed to the unreacted pendant alkyl groups of branched PEI, the additional thermal stability of the polymeric film above 400 °C is due to excessive crosslinking of amine groups to form a network structure.

Diffusion of organic molecules across the film

To demonstrate the diffusion of molecules through the film, we carried out experiment with a pH indicator, phenolphthalein. As expected, in a blank experiment without a film at the interface, phenolphthalein diffuses to aqueous layer within 1–2 minutes after its addition to organic layer. In a typical experiment, PU films with different thickness (80 nm to 290 nm) were prepared from PEI/HMDI at the CHCl₃/water interface. An appropriate volume of phenolphthalein stock solution was injected into the chloroform layer and an equal volume of chloroform was removed simultaneously to reduce pressure on the PU film (see Fig. S5, ESI†). The transport of phenolphthalein across PEI/HMDI PU film of 80 nm thickness towards alkaline aqueous solution starts to occur within 2 minutes and gradually the pink color intensifies with time (Fig. S6, ESI†). By using a thick PU film (200–290 nm), the diffusion rate was reduced and pink color begins to appear in the aqueous layer after 30 minutes. The transport across the membrane was quantified by calculating the phenolphthalein concentration in the aqueous layer by measuring the absorbance at λ_max = 552 nm (Fig. S7, ESI†). The calculated dye concentration was plotted against diffusion time for each film with varying thickness from 80–290 nm (Fig. 6A). As expected, increase in thickness of the film is expected to reduce porosity, thereby decrease the permeation of organic molecules.

Fig. 6 Changes in the concentration of phenolphthalein in the aqueous layer (as calculated by its absorbance at 552 nm) with increase in time and transported across (A) PEI film and (B) DETA film.

Similarly, diffusion of phenolphthalein across the PU membrane formed from DETA/HMDI at the liquid–liquid interface was also investigated. The experiment was repeated with PU films with an average thickness of 50 nm and 120 nm. Aliquots of aqueous solution were analyzed periodically using UV spectrofluorometer for monitoring the transport of the dye (Fig. 6B and S8, ESI†).

The PU film formed in 1 h is usually thinner than the ones formed after 24 h. As observed in the case of PEI/HMDI film, the diffusion of phenolphthalein through the membrane is inversely proportional to the thickness of the membrane. The diffusion of phenolphthalein through a thick membrane is limited and the aqueous amine layer appeared colorless after 60 minutes. The diffusion rate constants were deduced from the slopes of the curves as shown in Table 1. The rate of diffusion decreases with increase in the film thickness.

Table 1 Diffusion rate constants for the phenolphthalein transport across PEI–PU and DETA–PU films with varying film thickness

PEI film		DETA film
Thickness	k (μM min^−1)	Thickness	k (μM min^−1)
80 nm	5.40 × 10^−3	50 nm	3.80 × 10^−4
200 nm	2.10 × 10^−3	120 nm	1.19 × 10^−5
290 nm	9.66 × 10^−4

---

It is noted that the rate constants for PU films from PEI/HMDI are higher than those form DETA/HMDI. This is simply due to the differential transport of phenolphthalein across the PU film (Fig. 7).

Fig. 7 Schematic representation of the smooth (left side, PEI/HMDI) and defect (right side, DETA/HMDA) PU film and observed transport of phenolphthalein.

In case of PEI/HMDI film, a concentration gradient driven smooth diffusion of phenolphthalein from organic phase to the aqueous phase was observed, whereas in the case of DETA/HMDI film, the solubility driven transport of amines through the holes and tubules dominate and reaction occurred on the organic phase, especially at the tip of the tubules.

After exposure of the DETA–PU films to phenolphthalein, a shift of the urea C=O_str peak from 1633 cm⁻¹ to 1607 cm⁻¹ and a shift of the 1560 cm⁻¹ amide peak to 1512 cm⁻¹ are observed in the FT-IR spectra (Fig. 8B). Intensity of ring lactone C=O_str peak 1735 cm⁻¹ increases with the increasing concentration of phenolphthalein adsorbed on the surface of PU film. These observations show a strong interaction of phenolphthalein with the urea groups of DETA–PU film.

Fig. 8 FTIR spectra of PU films prepared via pre-adsorption of phenolphthalein at different concentrations from chloroform. The spectra were recorded by using the disk obtained from grinding the PU film with KBr powder. Spectra from PU–PEI (A) and PU–DETA (B) films are given with pure film (i) and film adsorbed with phenolphthalein at different concentrations, 0.1 mM (ii), 1 mM (iii) and 10 mM (iv).

Change in morphology of the films after exposure to phenolphthalein was also examined using SEM (Fig. 9). The morphology of the PU film from PEI/HMDI did not change significantly, but appeared smooth, except for a few lines visible on the surface (Fig. 9A–D). However, in the case of DETA, the diffusion of the amines from aqueous layer to the organic layer occurs through the number of holes and tubules on the membrane, followed by the reaction with diisocyanates and phenolphthalein in the organic phase (Fig. 9E–H). This led to the formation of needle shaped precipitates at the tip of the tubules on the organic side of the membrane (Fig. 9F) and prevents further diffusion of amines from the aqueous phase. This is also consistent with the IR data discussed above (Fig. 8B).

Fig. 9 FESEM micrographs of PU film from PEI/HMDI (A–D) and DETA/HMDI (E–H) before (A, C, E and G) and after (B, D, F and H) exposure to phenolphthalein dye in organic and aqueous phase.

Adsorption of fluorescent dyes via hydrogen bonding

The inter-chain hydrogen bonding contributes significantly to the morphology and 3-dimensional structure of the polymers. The availability of a large number of urea moieties on PEI/HMDI PU films provide multiple hydrogen bonding sites. In order to demonstrate adsorption capacities via hydrogen bonding, two fluorescent dyes with pendant carboxylic acid groups (Scheme 2) were selected and corresponding esters were used as controls. While 9-ACA showed emission maxima located at 416, 436 and 466 nm (λ_ex = 340 nm), the emission maxima of 1-PCA were seen at 382 and 402 nm with an intense broad peak centered at 450 nm (λ_ex = 340 nm).

The emission spectra of the dyes adsorbed on thin films at different time intervals (1, 3 and 5 h) are shown in Fig. 10A and B. Films with carboxylic acids showed higher emission intensities than the films exposed to ethyl esters (PU–ACEt and PU–PCEt), indicating strong adsorption of acidic compounds via hydrogen bonding between carboxylic acid and urea groups (Fig. 10C).

Fig. 10 Fluorescence emission spectra of PU films after adsorption of 9-ACA and 9-ACEt (A), 1-PCA and 1-PCEt (B) at different time intervals (1–5 h). The spectra of the thin films were obtained after excitation at λ = 340 nm. (C) Schematic diagram of the adsorption of ACA molecules on PU film via hydrogen bond formation between urea and carboxylic acid groups.

FTIR analysis further confirms the adsorption of 1-PCA and 9-ACA on the film (Fig. 11). The C=O stretching peaks corresponding to 9-ACA and 1-PCA are observed at 1681 and 1677 cm⁻¹, respectively and the carbonyl stretching of urea at 1633 cm⁻¹, however, no significant shift in the C=O stretching or NH-bending frequencies of urea was observed after adsorption. It is conceivable that only the surface urea groups are involved in hydrogen bonding with dyes, which is small in number as compared to the urea groups in the bulk of the PU film.

Fig. 11 FTIR spectra of PEI/HMDI PU thin film before and after adsorption of 9-ACA (A) and 1-PCA (B).

Conclusion

A convenient method for the synthesis of ordered PU films at the liquid–liquid interface using polyamines and diisocyanates was discussed. Thin films formed from PEI exhibited significant morphological differences as compared to those formed from small molecular amines e.g. DETA and TREN. Role of experimental conditions such as mole ratio of monomers and time of polymerization on the PU film thickness was investigated. PU films from PEI/HMDI showed smooth films, while PU films from DETA and TREN showed spherical and tube – type structures on the film surface due to diffusion of amines from aqueous to organic phase, followed by reaction with HMDI. Effect of film thickness on the diffusion of an organic molecule from organic to aqueous phase was demonstrated by the permeation of phenolphthalein dye across the film. PU films formed from PEI/HMDI showed smooth diffusion of phenolphthalein, whereas film form DETA/HMDI showed low rate of transport as a consequence of dye–amine salt formation. Similarly, the surface functional groups were used for the adsorption of organic molecules, 9-ACA and 1-PCA, through hydrogen bonding. In addition, by tuning the film thickness, it is possible to use the membrane for selective transport of various organic molecules across the liquid–liquid interface.

Acknowledgements

The authors thank the technical support from Department of Chemistry, National University of Singapore. R. D. and S. D. thank National University of Singapore for graduate scholarship.

References

1. F. Khelifa, S. Ershov, M.-E. Druart, Y. Habibi, D. Chicot, M.-G. Olivier, R. Snyders and P. Dubois, J. Mater. Chem. A, 2015, 3, 15977.
2. (a) K. P. Lee, T. C. Arnot and D. Mattia, J. Membr. Sci., 2011, 370, 1; (b) R. J. Petersen, J. Membr. Sci., 1993, 83, 81; (c) S. S. Dhumal, S. J. Wagh and A. K. Suresh, J. Membr. Sci., 2008, 325, 758.
3. K. P. Lee, J. Zheng, G. Bargeman, A. J. B. Kemperman and N. E. Benes, J. Membr. Sci., 2015, 478, 75.
4. M. Masahiro, K. Yasuko, M. Masahiro and I. Kenji, Jpn. J. Appl. Phys., 2015, 54, 04DK13.
5. (a) M. Pontié, H. Dach, J. Leparc, M. Hafsi and A. Lhassani, Desalination, 2008, 221, 174; (b) W. Gao, F. She, J. Zhang, L. F. Dumée, L. He, P. D. Hodgson and L. Kong, J. Membr. Sci., 2015, 487, 32.
6. (a) H. Zhou, M. F. Toney and S. F. Bent, Macromolecules, 2013, 46, 5638; (b) H. Zhou and S. F. Bent, ACS Appl. Mater. Interfaces, 2011, 3, 505; (c) C. Prasittichai, H. Zhou and S. F. Bent, ACS Appl. Mater. Interfaces, 2013, 5, 13391; (d) P. W. Loscutoff, H. Zhou, S. B. Clendenning and S. F. Bent, ACS Nano, 2010, 4, 331.
7. Y. Fu, B. Li, Y.-B. Jiang, D. R. Dunphy, A. Tsai, S.-Y. Tam, H. Fan, H. Zhang, D. Rogers, S. Rempe, P. Atanassov, J. L. Cecchi and C. J. Brinker, J. Am. Chem. Soc., 2014, 136, 15821.
8. (a) B. Vonhören, O. Roling, K. De Bruycker, R. Calvo, F. E. Du Prez and B. J. Ravoo, ACS Macro Lett., 2015, 4, 331; (b) J. Wang, J. Li, N. Li, X. Guo, L. He, X. Cao, W. Zhang, R. He, Z. Qian, Y. Cao and Y. Chen, Chem. Mater., 2015, 27, 2439; (c) M. Kim, M. Byeon, J.-S. Bae, S.-Y. Moon, G. Yu, K. Shin, F. Basarir, T.-H. Yoon and J.-W. Park, Macromolecules, 2011, 44, 7092.
9. H. Takeshi, I. Masayuki, T. Yoshikazu, F. Eiichi, S. Yoshiichi, K. Masa-aki and I. Yoshio, Jpn. J. Appl. Phys., 1995, 34, 395A.
10. (a) B. Waghmode, S. Patil, M. Jahagirdar, V. Patil, R. Waichal, D. Malkhede, S. Sathaye and K. Patil, Colloid Polym. Sci., 2014, 292, 1079; (b) K. Wójcik and M. Grzeszczuk, J. Solid State Electrochem., 2015, 19, 1293; (c) M. Liu, Y. Geng, Q. Wang, Y.-I. Lee, J. Hao and H.-G. Liu, RSC Adv., 2015, 5, 4334.
11. (a) J. Park, M. Goh and K. Akagi, Macromolecules, 2014, 47, 2784; (b) S. S. Dhumal and A. K. Suresh, Polymer, 2010, 51, 1176; (c) S. K. Yadav, A. K. Suresh and K. C. Khilar, AIChE J., 1990, 36, 431; (d) Y. Song, P. Sun, L. L. Henry and B. Sun, J. Membr. Sci., 2005, 251, 67.
12. M. W. Patchan, B. W. Fuller, L. M. Baird, P. K. Gong, E. C. Walter, B. J. Vidmar, I. Kyei, Z. Xia and J. J. Benkoski, ACS Appl. Mater. Interfaces, 2015, 7, 7315.
13. Z. Tang, D. Sun, D. Yang, B. Guo, L. Zhang and D. Jia, Compos. Sci. Technol., 2013, 75, 15.
14. S. Natour and R. Abu-Reziq, RSC Adv., 2014, 4, 48299.
15. X. Ouyang, X. Huang, Q. Pan, C. Zuo, C. Huang, X. Yang and Y. Zhao, J. Dent., 2011, 39, 825.
16. (a) J. Shim and D. Mohr, Int. J. Impact Eng., 2009, 36, 1116; (b) H. Han, S. Li, X. Zhu, X. Jiang and X. Z. Kong, RSC Adv., 2014, 4, 33520; (c) S. Das, I. Yilgor, E. Yilgor, B. Inci, O. Tezgel, F. L. Beyer and G. L. Wilkes, Polymer, 2007, 48, 290.
17. (a) D. Fragiadakis, R. Gamache, R. B. Bogoslovov and C. M. Roland, Polymer, 2010, 51, 178; (b) N. Tsutsumi, Y. Okabe and W. Sakai, Macromolecules, 1999, 32, 3249; (c) A. Kim, M. A. Filler, S. Kim and S. F. Bent, J. Am. Chem. Soc., 2005, 127, 6123; (d) T. Yoshikazu, I. Masayuki and F. Eiichi, Jpn. J. Appl. Phys., 1989, 28, L2245.
18. I. Polenz, S. S. Datta and D. A. Weitz, Langmuir, 2014, 30, 13405.
19. (a) S. K. Yadav, N. Ron, D. Chandrasekharam, K. C. Khilar, A. K. Suresh and V. M. Nadkarni, J. Macromol. Sci., Part B: Phys., 1996, 35, 807; (b) S. K. Yadav, K. C. Khilar and A. K. Suresh, AIChE J., 1996, 42, 2616; (c) M. Kubo, Y. Harada, T. Kawakatsu and T. Yonemoto, J. Chem. Eng. Jpn., 2001, 34, 1506.
20. S. V. Joshi and A. V. Rao, J. Appl. Polym. Sci., 1991, 42, 1773.
21. L. Zhu, C.-Y. Huang, Y. H. Patel, J. Wu and S. V. Malhotra, Macromol. Rapid Commun., 2006, 27, 1306.
22. P. Nava, M. Cecchini, S. Chirico, H. Gordon, S. Morley, D. Manor and J. Atkinson, Bioorg. Med. Chem., 2006, 14, 3721.
23. V. Freger, Langmuir, 2005, 21, 1884.
24. S. J. Wagh, S. S. Dhumal and A. K. Suresh, J. Membr. Sci., 2009, 328, 246.
25. (a) Y.-T. Chern and B.-S. Wu, J. Appl. Polym. Sci., 1996, 61, 1853; (b) Y.-T. Chern, B.-S. Wu and C.-M. Huang, Polym. Bull., 1996, 36, 609.
26. (a) J. Ji, J. M. Dickson, R. F. Childs and B. E. McCarry, Macromolecules, 2000, 33, 624; (b) S. K. Karode, S. S. Kulkarni, A. K. Suresh and R. A. Mashelkar, Chem. Eng. Sci., 1997, 52, 3243.
27. M. Jahanshahi, A. Rahimpour and M. Peyravi, Desalination, 2010, 257, 129.

---

Footnote

† Electronic supplementary information (ESI) available: Additional data (SEM, AFM micrographs and UV spectra) on characterization of the PU films is available. See DOI: 10.1039/c5ra27108h

---